Accutane (Isotretinoin) Bone Health and Density Impact

What the label lists for bone

The Absorica and Absorica LD label carries a Warnings and Precautions heading for musculoskeletal abnormalities: arthralgias, back pain, decreases in bone mineral density, and premature epiphyseal closure [1]. Its adverse reaction list adds skeletal hyperostosis and calcification of tendons and ligaments to the bone density and growth plate entries [1]. The older-format Accutane labeling has a skeletal warnings section covering bone mineral density, hyperostosis, and premature epiphyseal closure, and lists the same reactions [2].

Back pain, arthralgia, and musculoskeletal discomfort are among the adverse reactions the Absorica label lists as occurring in at least 5% of patients [1]. The label says the listed reactions came from clinical trials or voluntary reports from populations of uncertain size, so frequency cannot always be reliably estimated and a causal relationship to the drug cannot always be established [1]. On duration, the label says: 'Long-term use of ABSORICA/ABSORICA LD, even in low dosages, has not been studied, and is not recommended. The effect of long-term use of ABSORICA/ABSORICA LD on bone loss is unknown' [1].

Two prospective studies at acne doses

A multicenter, open-label study by DiGiovanna and colleagues measured bone mineral density by dual energy x-ray absorptiometry in 217 adolescents (81 girls), aged 12 to 17, with severe recalcitrant nodular acne, at baseline and at the end of one course of about 1 mg/kg/day taken twice daily with food for 16 to 20 weeks [3]. Mean change was +1.4% at the lumbar spine (range -4.9% to +12.3%) and -0.26% at the total hip (range -11.3% to +15.0%) [3]. The authors concluded there was no clinically significant effect, and hyperostosis was not observed in any patient [3].

Leachman and colleagues studied young men aged 17 to 25 receiving isotretinoin for cystic acne, compared with healthy volunteers aged 19 to 26 [4]. Before any treatment, mean bone density in the acne patients was lower at all measured sites (spine, femoral neck, Ward triangle) and more variable at the spine than in controls. After 6 months at 1 mg/kg, density at the Ward triangle fell a mean of 4.4% (P = .03); four patients lost more than 9% at that site [4]. The difference between groups was significant at the Ward triangle (P = .04) but not at the other sites, and calcium metabolism measurements did not change in either group [4]. The authors concluded that bone loss without measurable change in calcium metabolism is likely a direct effect of retinoids on bone [4].

The two studies differ in population, duration, and the sites that changed [4][3].

Growth plates and teenagers

Premature epiphyseal closure appears on the label as an adverse reaction [1][2]. The clearest published description is a single case report: a boy with epidermolytic hyperkeratosis treated systemically with 13-cis-retinoic acid for 4 and a half years developed knee pain at age 10 and a half, with radiographic partial closure of the proximal epiphysis of the right tibia; the authors noted similar changes described in people ingesting excessive amounts of vitamin A [5]. That child was treated for a keratinization disorder, not acne, and for years rather than months.

The label also notes that the optimal interval before a second course has not been defined for patients who have not completed skeletal growth, and it does not recommend a second course before a 2-month waiting period [1]. In the 16- to 20-week adolescent study, no hyperostosis was observed, and the mean changes in spine and hip bone density were not clinically significant [3].

Hyperostosis and long courses

Skeletal hyperostosis and calcification of tendons and ligaments are listed reactions [1][2]. No hyperostosis was observed in the adolescent acne study [3]. The background section of the DiGiovanna paper notes that adverse bone changes have been reported in patients on high-dose, long-term isotretinoin therapy, lasting several years, for disorders of cornification, which is a different use from the approved acne indication [3][1].

Other drugs that matter for bone

The Accutane label's drug interaction section addresses two bone-relevant combinations. It says systemic corticosteroids are known to cause osteoporosis and that no formal clinical studies have assessed an interactive effect on bone loss between systemic corticosteroids and isotretinoin, so caution should be exercised when the drugs are used together [2]. It says phenytoin is known to cause osteomalacia, and again no formal studies have assessed an interactive effect on bone loss with isotretinoin [2].

For tetracyclines, the label says to avoid concomitant use because isotretinoin has been associated with intracranial hypertension (pseudotumor cerebri), some cases involving tetracyclines [1][2].

For vitamin A, the Absorica label says to avoid supplements containing vitamin A, because isotretinoin is closely related to vitamin A and combined use may lead to isotretinoin-associated adverse reactions [1].

What animal and cell studies show

In rats pair-fed a high vitamin A intake, body weight and femur length were unchanged compared with controls, but bone formation and mineral apposition rate at the femur diaphysis were roughly halved [6]. In cultured human osteoblasts and a mouse preosteoblast cell line, retinoic acid suppressed mineralization through retinoic acid receptors, independent of any effect on cell proliferation, and drastically reduced alkaline phosphatase and osteocalcin, along with the transcription factors Runx2 and Osterix and the osteoblast protein Phex [6]. That work used vitamin A and retinoic acid, not isotretinoin, and its relevance to a 15- to 20-week acne course is not established.

Cumulative dose and bone outcomes

The label does not state a cumulative mg/kg target. It gives 0.5 to 1 mg/kg/day in two divided doses for 15 to 20 weeks, with up to 2 mg/kg/day for adults with very severe disease, and notes that if the nodule count falls by more than 70% before 15 to 20 weeks the drug may be stopped [1][2].

A 2026 single-centre, single-blind randomized trial compared cumulative doses of 120 and 150 mg/kg in moderate-to-severe cystic acne. Lesion counts and severity improved similarly in both groups (all p > 0.20), 12-month relapse was 26.7% with 120 mg/kg and 32.3% with 150 mg/kg (p = 0.619), and adverse events were mild and comparable [7]. The outcomes listed were acne severity, lesion counts, relapse, and scarring; bone density was not among them [7].

The 1984 Strauss dose-response trial enrolled 150 patients with treatment-resistant nodulocystic acne in three equal groups at 0.1, 0.5, and 1.0 mg/kg/day. All three doses produced a highly significant response with no significant difference between them, but 42% of the 0.1 mg/kg/day group needed retreatment [8]. Its listed outcomes were clinical response, side effects, laboratory abnormalities, and remission duration [8].